A powerful command-line tool and library to find typos in your JavaScript, TypeScript, JSX, and TSX source code files by analyzing string literals using AST parsing and spell-checking with nspell.
- File discovery: Uses fast-glob to find all JS/TS source files.
- Parses JS/TS/JSX/TSX files and analyzes string literals only.
- Uses robust AST parsing (
@typescript-eslint/typescript-estree) to precisely locate strings and their line numbers. - Uses
nspellwith English dictionary for spell checking. - Builds a project-specific dictionary of known correct words for context-aware validation.
- Supports a configurable whitelist of allowed words and technical terms.
- Filters out acronyms, short words, and known technical terms from typo results.
- Suggests corrections for detected typos.
- Displays results in a clear CLI table with file, line, typo word, and suggestions.
- Usable as both a standalone CLI tool or as a programmatic library.
npm install typo-checkerOR with yarn
yarn add typo-checkerIn your project’s package.json, add:
"scripts": {
"typo-checker": "typo-checker"
}You can customize the whitelist of allowed words (to reduce false positives) by creating a typo-checker.config.json in your project root:
{
"whitelist": [
"React",
"Redux",
"NodeJS",
"GraphQL",
"uuid",
"TSConfig"
]
}
Alternatively, add it inside your package.json:
{
"typoChecker": {
"whitelist": [
"React",
"Redux",
"NodeJS"
]
}
}
If no typos are found:
- Built with TypeScript.
- Uses ES modules.
- Dependencies:
fast-glob,chalk,cli-table3, nspell,dictionary-en,@typescript-eslint/typescript-estreenatural.
- Parses JS/TS/JSX/TSX files and analyzes string literals only.
- Only checks string literals inside source code — does not analyze comments or identifiers.
- Works best for English text.
- Whitelist is critical to avoid noise from domain-specific words.
- Acronyms and short words are ignored.
- Handles US/UK spelling variants gracefully, avoiding false positives.
Feel free to open issues or PRs for new utility suggestions or improvements.
MIT